Pixelation

[ptc-1368288]

In case the search tool will be restored,
this important work is posted !searchable!

jmG

[ptc-1368288]

... Quite an "ad absurdum" proof that resample2D(,,,) is worth what it worths. "Deceptively simple or simply deceptive [Richard]". I love that quote ! Still, very interesting and educative, though my A is a bit underdesigned. "Sur le fond", isn't pretentious that by plugging a series of 0's in an image it would just add pixels, valuable pixels ? Preaching me more is like preaching on the "wall of lamentations", preaching in the blue.



jmG

[ptc-1368288]

...
Pixelation is by "zoom(,,,), and that's it.
All the theories of padding Fourier is now in the coffin. Fourier is essentially "exponential maths", as well as the Gaussian spot. Remembering the mass of properties of the "exponential", not surprising some parenthood match for the spot, but this property of padding seems unique vis Fourier 0's padding. Years ago [6 or +] I have demonstrated the difficulty of fitting "Gaussian like" and exponentials on their "reflexive" properties. More recently, the plasma model is one of this most difficult fitting case, though well done for the most various fits. Those interested in pixelation should concentrate on getting the coding of "zoom(,,,)" for applications outside of the Mathcad Improc. Whether particular to the Mathcad Improc or a universal coding, the matter is that it's the function that "pixelate". End of story.

jmG

[ptc-1368288]

...

http://www.magictoolbox.com/magiczoom/

[Jbryant61]

Hi Jean. I have taken a look at your pad function.
What i have said before that Lou's Resample fits my data (at least in the background) much better than Virendra's, I have noticed it isn't fitting the peak of the data too good (maybe I have done something in the worksheet.)

I wanted to take a look at your Pad function.

I want to cut slices thru' my original data, and then find the corresponding slice thru your pad function and plot both as linescans to see how good each on it. I want a variable n tod efine the slice pixel number, varying from 0 24 (my original data has 25 rows & columns.

Im not sure how to convert your pad image to the same linescan as the original data?

Jason

[ptc-1368288]

>Hi Jean. I have taken a look at your pad function.
What i have said before that Lou's Resample fits my data (at least in the background) much better than Virendra's, I have noticed it isn't fitting the peak of the data too good (maybe I have done something in the worksheet.)

I wanted to take a look at your Pad function.

I want to cut slices thru' my original data, and then find the corresponding slice thru your pad function and plot both as linescans to see how good each on it. I want a variable n to define the slice pixel number, varying from 0 24 (my original data has 25 rows & columns.

Im not sure how to convert your pad image to the same linescan as the original data?

Jason<<br> _____________________________

The pad function is the 1/1 resample2D. Simpler and more visual. No problem to visualize two superposed slices, but in reference to their respective "max", they don't coincide by quite a few pixels, in fact by � the pixelation factor. In other words, resample2D offsets in both directions by that factor.

Jean